do body size and diet breadth affect partitioning of species diversity? a test with forest...

© 2006 The Authors DOI: 10.1111/j.1366-9516.2006.00206.x

91

Journal compilation © 2006 Blackwell Publishing Ltd www.blackwellpublishing.com/ddi

Diversity and Distributions, (Diversity Distrib.)

(2006)

12

, 91–99

BIODIVERSITYRESEARCH

ABSTRACT

Two of the major themes resulting from recent macroecological research are thecentral roles that body size and niche breadth may play as determinants of speciesgeographical distribution. Unanswered questions, however, linger regarding howsimilarities in body size or niche breadth affect the allocation of

α

- and

β

-diversityacross spatial scales. Using data on moth diversity in the eastern deciduous forest ofNorth America, we tested the predictions that smaller-bodied and diet-restrictedspecies would have lower levels of

α

-diversity within forest stands and greater

β

-diversity at higher sampling scales compared to larger or more generalist species.Moths were sampled using a nested sampling design consisting of three hierarchicallevels: 20 forest stands, 5 sites and 3 ecoregions. Body size for 492 species was estimatedas mean forewing length, and diet breadth was assessed from the published literature.Moth species were then classified according to body size (small or large) or dietbreadth (generalist or restricted), and partitioning was conducted on each group.Diversity partitions for large- and small-bodied species yielded similar patterns.When observed diversity components differed from those derived from our nullmodel, a consistent pattern was observed:

α

-diversity was greater than expected,

β

-diversity among forest stands was less than expected, and

β

-diversity among sites andecoregions was higher than expected. In contrast, diet-restricted moths contributedsignificantly less to stand-level

α

-diversity than generalist feeders. Furthermore,specialists contributed to a greater proportion of

β

-diversity across scales comparedto generalist moths. Because absolute measures of

β

-diversity among stands weregreater for generalists than for restricted feeders, we suggest that regional

β

-diversityof forest moths may be influenced by several possible factors: intraspecific aggregationof diet-restricted species, local fluctuations in population size of eruptive generalistsand small geographical distributions of generalist moths than predicted by thegeographical extent of putative host plants

Keywords

Additive diversity partition,

β

-diversity, macroecology, moths, regional diversity,

species composition.

INTRODUCTION

Macroecology is primarily concerned with detecting emergent

properties of ecological communities and finding parsimonious

explanations for patterns of species distribution and abundance

at broad spatial scales (Brown, 1995). Two of the major themes

resulting from recent macroecological research are the central

roles that both body size and niche breadth play as determinants of

species bionomics and geographical distribution (e.g. Blackburn

& Gaston, 2003). To more fully understand the organization of

ecological communities, it is necessary to test whether differences

in species traits (e.g. body size, niche specialization) lead to non-

random differences in the allocation of species diversity across

spatial scales (e.g. Summerville

et al

., 2003). Veech

et al

. (2003)

offered a preliminary assessment of factors influencing diversity

components in space by demonstrating that significant intra-

specific aggregation is correlated with low local species diversity

(

α

-diversity) and high species turnover (

β

-diversity) within a

region. The recently derived analytical approach to diversity par-

titioning (see Crist

et al

., 2003) may further provide a framework

1

Department of Environmental Science and

Policy, Drake University, Des Moines, Iowa

50311 USA;

2

Department of Biology, University

of Northern Colorado, Greeley, Colorado 80639

USA; and

3

Department of Zoology, Miami

University, Oxford, Ohio 45056 USA

*Correspondence: Keith S. Summerville, Department of Environmental Science and Policy, Drake University, Des Moines, Iowa 50311 USA. Tel.: (515) 271-2265; Fax: (515) 271-3702; E-mail: [email protected]

Blackwell Publishing, Ltd.

Do body size and diet breadth affect partitioning of species diversity? A test with forest Lepidoptera

Keith S. Summerville

1

*, Tiffany D. Wilson

1

, Joseph A. Veech

2

and

Thomas O. Crist

3

K. S. Summerville

et al.

92

© 2006 The Authors

Diversity and Distributions

,

12

, 91–99, Journal compilation © 2006 Blackwell Publishing Ltd

to examine whether body size or niche breadth leads to similar

patterns in the allocation of

α

- and

β

-diversity components in

space described in Veech

et al

. (2003).

Body size is both easy to measure and is correlated with species

dispersal ability (Gardezi & da Silva, 1999; Bowman

et al

., 2002;

Brändle

et al

., 2002). The relationship between body size and

geographical range size, however, is not necessarily linear. Brown

(1995) and Blackburn and Gaston (2003) suggest that species

with either large or very small body sizes tend to possess larger

home ranges, but species with intermediate body sizes tend to

show a positive correlation between body size and geographical

range (the complete relationship this thus triangular). As postu-

lated by Brändle

et al

. (2002), the body size–geographical range

hypothesis states that when large-bodied, non-parasitic organisms

are more vagile, they should be present at a larger number of sites

within their geographical range or over a larger total area than

smaller-bodied organisms (see also Swihart

et al

., 2003). Given

the potential positive correlations between body size and dispersal

ability, body size may also affect the partitioning of

α

- and

β

-

diversity components from a regional species pool (Gaston &

Reavey, 1989). Larger-bodied species should have a relatively high

α

-diversity within ecological communities, while smaller-bodied

species, which may be more limited in geographical distribution,

should have significantly low

α

-diversity and greater

β

-diversity

among communities within a region. In contrast, Gaston’s (1988)

study of local and regional patterns in moth population dynamics

suggests that an opposite pattern may actually be observed: smaller-

bodied species may be the most widely distributed although

populations of such species can be highly variable and experience

frequent local extinction (see also Blackburn & Gaston, 1996).

The potential effects of niche breadth on community organiza-

tion have been studied most extensively using diet specialization,

with studies demonstrating that species known to feed on a

greater range of resources possess greater potential fecundity

(Spitzer

et al

., 1984; Hunter, 1991), larger geographical distribution

(Gaston, 1988; Roy

et al

., 1998), and perhaps greater susceptibility

to environmental stochasticity (Rajmánek & Spitzer, 1982; Spitzer

et al

., 1984; Hunter, 1991). These patterns remain somewhat

equivocal, as other studies have found no relationship or even the

opposite relationships between a species diet and the above

variables (Redfearn & Pimm, 1988; Lep

Í

et al

., 1998). It remains

plausible, however, that dietary generalists have a greater likeli-

hood of having their resource needs met in a greater number of

habitats, whereas specialists should be highly restricted to a

smaller subset of habitat patches within the greater landscape

(Roy

et al

., 1998; Summerville

et al

., 2002; Swihart

et al

., 2003).

In terms of hierarchical patterns of species diversity, generalists

should have higher levels of

α

-diversity within communities

compared to specialists, which should have higher levels of

β

-

diversity among communities (e.g. Tscharntke

et al

., 2002).

The goal of this study was to determine if body size and diet

breadth affect

α

- and

β

-diversity across a number of sampling

scales. Diversity partitioning is increasingly being used to examine

patterns in

α

- and

β

-diversity within a regional species pool

(Veech

et al

., 2002; Crist

et al

., 2003). First, we tested the prediction

that large-bodied moths would display greater levels of

α

-diversity

and lower levels of

β

-diversity within communities compared to

smaller-bodied species. Second, we tested the prediction that

moth species with a broader diet breadth would have higher

α

-

diversity and lower

β

-diversity compared to specialists. Both tests

were conducted using data on moth communities from the eastern

deciduous forest in Ohio, USA. The Lepidoptera were well-suited

to this study because data regarding both body size and diet

breadth are well-documented in the literature (e.g. Summerville

& Crist, 2002).

METHODS

Study sites and sampling design

We used a nested design to sample Lepidoptera from stands

of eastern deciduous forest in Ohio, USA. Three hierarchical

levels comprised the design, listed from broadest to smallest:

ecoregions, forest sites and forest stands (see Summerville

et al

.,

2003). Ecoregions differed in their glacial history, topographical

heterogeneity, soil types and floristic composition (Fig. 1). The

forests of the North Central Tillplain (NCT) are dominated by

American beech (

Fagus grandifolia

Ehrh.) and sugar maple

(

Acer saccharum

Marsh.) (Braun, 1961). Species such as white

oak (

Quercus alba

L.), red oak (

Quercus rubra

L.), slippery elm

(

Ulmus rubra

Muhl.) and several ashes (

Fraxinus

spp.) are also

important canopy species (Greller, 1988). Land use in the NCT is

predominantly agricultural, since glacial scouring has created a

flat topography (ridges are separated by shallow, sloping flood-

plains) and productive soils. In contrast, the Western Allegheny

Plateau (WAP) and the Interior Low Plateau (ILP) largely escaped

Pleistocene glaciation. The WAP is characterized by acidic, less

productive soils and a topography characterized by steep ridges and

long, narrow drainages. In the WAP, xeric aspects are dominated

by chestnut oak (

Quercus prinus

L.) and hickories (

Carya

spp.),

Figure 1 Map of Ohio showing ecoregions and study site locations. Dark lines represent ecoregional boundaries; � represents study sites. Map adapted from The Nature Conservancy (1999). Study sites are listed clockwise beginning from the top of the Western Allegheny Plateau: Clear Creek MetroPark, Shawnee State Forest, Edge of Appalachia, Caesar Creek State Park and Hueston Woods State Park.

Body size, diet breadth and diversity partitioning

© 2006 The Authors

93


,

12


while mesic areas contain a more diverse assemblage of trees

including American beech, tulip poplar (

Liriodendron tulipifera

L.), basswood (

Tilia americana

L.) and eastern hemlock (

Tsuga

Canadensis

L. (Carr.)) (Greller, 1988). The ILP of southern Ohio is

similar to the WAP in its glacial history and variable topography.

Portions of the ILP, however, are underlain by dolomitic

limestone and have more alkaline soils that tend to support a

greater overall diversity of vegetation.

Our experimental design nested two sites within the NCT and

the WAP (Fig. 1). Hueston Woods State Park (HWSP; Preble

Co., OH) and Caesar Creek State Park (CACR; Warren Co., OH)

occurred in the glaciated NCT, whereas Clear Creek MetroPark

(CLCR; Hocking Co., OH) and Vastine Wilderness Area (VAST;

Scioto Co., OH) occurred in the unglaciated WAP ecoregion. We

included a fifth site in our study, Edge of Appalachia nature

preserve (EDGE; Adams Co., OH) that falls within the ILP (Fig. 1).

Within each site, we selected four forest stands (

c

. 1 ha) that

represent typical mesic and xeric aspects and were separated by

a minimum distance of 250 m from other stands as well as other

ecotones. Because the EDGE falls near the transition zone

between the WAP and the ILP, we selected stands for sampling

at the EDGE that occurred on geological formations more

characteristic of the WAP. Mesic stands contained woody and

herbaceous flora similar to bottomlands at Shawnee State Forest

(Summerville

et al.

, unpublished data). A previous study dem-

onstrated that the lepidopteran fauna of EDGE is closely affili-

ated with that of the WAP ecoregions (Summerville

et al

., 2003).

Lepidoptera sampling

Within each forest stand, we sampled Lepidoptera using a single

12-W universal blacklight trap (BioQuip Products, Inc., Rancho

Dominguez, CA) powered by a 12-V (26 amp h

−

1

) gel battery.

Blacklight traps are widely recognized to be the standard

technique for sampling moth communities, although the method is

biased toward collecting phototactic species. To avoid disruption

of the UV light by seedlings and shrubs in the understorey, we

positioned UV traps on platforms

c

. 1.5 m above the ground. We

sampled the moth communities of each forest stand during two

sampling periods over the summer of 2000: 15 May

−

1 June (‘early’)

and 29 July

−

8 August (‘late’). Sampling was seasonally stratified

because temporal variation has significant effects on lepidopteran

community structure, and our early and late sampling intervals

roughly correspond to the peaks in species richness for moths in

temperate forest systems (Thomas & Thomas, 1994). We operated

traps within each stand for two non-consecutive nights from

1930 to 0600 EDT during both early and late seasons (four nights

total per stand). Therefore, trapping accumulated 80 samples in

the early and the late seasons combined, representing

c

. 60% of

the moth fauna predicted to be present using the incidence-based

coverage (ICE) diversity estimator (see Longino

et al

., 2002).

Determining moth body size and diet breadth

We estimated body size as the mean of the maximum and

minimum forewing length (in cm), as reported in Forbes (1923),

Covell (1983) and other published monographs. Gaston and

Reavey (1989) and Hawkins and Lawton (1995) considered this

approach valid for large data sets, especially for insects. Using

this approach, body size could only be estimated for those species

we were able to identify. Summerville

et al

. (2003) collected a total

of 636 species of moths, of which 492 were identified to species

(Table 1). Some bias exists in the range of body sizes covered in

our data as very small-bodied moths are difficult to identify even

by experts (‘microlepidoptera’; e.g. Gracillaridae, Gelechioidea),

and published literature for species natural history is scant. Thus,

our data set contained fewer tiny moths (wingspan < 0.4 cm)

than likely occur in typical forest moth communities. Small

Table 1 Number of moth species and range in body sizes sampled from forest stands. Moth families are arranged in order of increasing species richness. Family names follow those in Hodges et al. (1983) with revisions according to Hodges (1999)

Family

Number of

moth species

Observed range

in body size

(min.–max. in cm)

Argyresthidae 1 1.35

Bucculatracidae 1 0.45

Coleophoridae 1 1.20

Drepanidae 1 2.95

Epiplemidae 1 2.00

Heliozelidae 1 0.75

Incurvariidae 1 1.25

Plutellidae 1 1.30

Pterophoridae 1 1.35

Sessiidae 1 2.15

Tineidae 1 1.40

Zygaenidae 1 2.30

Acrolophidae 2 2.85–3.0

Apetelodidae 2 3.70

Megalopigidae 2 3.00–3.25

Mimaloniidae 2 2.60–4.25

Yponmeutidae 2 1.85–2.40

Amphisbatidae 3 1.35–2.05

Gracillariidae 3 1.10–1.30

Lasiocampidae 3 3.00–3.90

Lymantriidae 4 2.65–3.90

Oecophoridae 4 0.45–1.90

Cosmopterigidae 5 0.55–1.75

Sphingidae 6 5.75–9.40

Elachistidae 9 0.70–1.80

Saturniidae 10 4.30–12.70

Gelechiidae 12 1.15–2.50

Limacodidae 14 2.00–2.95

Arctiidae 26 1.65–7.4

Notodontidae 28 2.90–5.20

Crambidae 30 0.70–3.35

Pyralidae 32 1.10–3.80

Tortricidae 53 1.00–2.60

Geometridae 84 1.30–5.20

Noctuidae 144 1.25–7.75

Total 492

K. S. Summerville

et al.

94

© 2006 The Authors


,

12


moths represented body sizes < 2.75 cm in wingspan; wingspan

of large moths was 2.75 cm (Fig. 2). Median body size for all

moths was 2.70 cm, so this classification resulted in only slight

differences in the total number of species within each group.

Small moths were mostly Tortricidae, Pyralidae and Crambidae

and Gelechiidae, while large moths were generally species of

Noctuidae, Geometridae, Arctiidae and Notodontidae (Table 1).

We used the data on diet breadth in Robinson

et al

. (2002) to

classify a total of 456 moth species as generalists, oligophagous or

specialist. Following Lep

Í

et al

. (1998) and Summerville and

Crist (2002), generalists were considered to be species known to

feed on hosts in a number of plant families, oligophages were

moths reported to feed on a number of host plant genera within

a single family, and specialists were species known only to feed

on plants within a single genus. This classification rendered

relatively few oligophagous moths, so we combined the oligo-

phagous and specialist species as a single diet-restricted group for

our data analyses (Summerville & Crist, 2002).

Analysis using additive partitioning of species diversity across scales

Diversity partitioning is emerging as a powerful tool to detect

changes in diversity and composition across sampling scales

(Lande, 1996; Wagner

et al

., 2000; Gering

et al

., 2002; Ouin

& Burel, 2002; Summerville

et al

., 2003). The formal analytical

basis of diversity partitioning has been well-developed elsewhere,

but briefly, the method measures total species diversity (

γ

) and

then partitions this measure into the additive components of

alpha- (

α

) and beta-diversity (

β

) (see Lande, 1996). We compared

observed

α

- and

β

-diversity at each hierarchical level of our

sampling design to expected values obtained through a random-

ization technique (Crist

et al

., 2003).

Within the context of our study,

α

- and

β

-diversity are defined

relative to a given level of sampling within the nested design.

Thus,

α

1

represents the mean diversity of moths within a forest

stand, and

β

1

represents the diversity among the 20 forest stands.

Because

α

-diversity in a given sampling level is simply the sum of

the

α

- and

β

-diversity at the next lowest level (Wagner

et al

.,

2000), the overall moth species diversity within the five sites in

our study can be expanded by the following formula:

α

2(sites)

=

α

1(stands)

+

β

1(stands)

. (1)

Similarly,

α

3(regions)

=

α2(sites) + β2(sites), and at the highest level,

the total diversity, γ = α3(regions) + β3(regions). By substitution, the

additive partition for our study is:

γ = α1(stands) + β1(stands) + β2(sites) + β3(regions). (2)

The total diversity can therefore be expressed as the propor-

tional contributions of diversity due to each level in the hierarchical

sampling design (see Summerville et al., 2003).

There is a multitude of ways to describe species diversity,

but two of the most commonly employed measures are species

richness and Simpson diversity (Magurran, 1988). Species

richness is simply a count of the number of species present within

a community, while the Simpson index is a measure of dominance

within a community as determined from the relative abundance

of species (Krebs, 1999). Simpson diversity was calculated as the

Gini coefficient: , where p is the relative abun-

dance of species i. Using these diversity indices, we conducted

additive partitions for the entire data set in both sampling periods

for a total of 16 partitions (two diversity metrics × two sampling

seasons × four moth groups) using a self-contained computer

program, (Crist et al., 2003).

For our analyses, we tested the diversity components from the

hierarchical sampling design against the null hypothesis that the

observed component of diversity at each level could have been

obtained by a random distribution of species within samples. In

effect, this hypothesis asks the question: for each class of body

size or diet breadth, is an observed component of diversity at a

given spatial level greater or less than one would expect by random

distribution of species in space? To test this hypothesis, we used a

reshuffling algorithm in to randomize samples (i.e.

trap catches within single forest stands or species sampled from

stands within forest sites) while conserving the observed species-

abundance and sample-size distributions. Samples at progres-

sively higher levels in our design were obtained by pooling the

appropriate samples at lower levels (just as done in our original

partition, Crist et al., 2003). For example, to test the significance

of observed β-diversity within a site, we randomized samples

from all the stands within a given ecoregion (Crist et al., 2003).

We repeated these randomizations 10,000 times to form a null

distribution of each α and β estimate (species richness and

Simpson diversity). Each of the level-specific estimates is then

compared to the appropriate null distribution. Statistical signi-

ficance is assessed by determining the proportion of null values

that are greater or less than our observed estimate. For example,

if 1 out of 10,000 null values is greater than the observed estimate,

then the probability of obtaining (by chance) an estimate greater

than the observed value is 0.0001. It is important to note that our

diversity-partitioning algorithm will allow us to compare diversity

components from different partitions in terms of whether one is

Figure 2 Distribution of body sizes for 492 moth species sampled in 2000. The dashed line breaks the distribution into two groups of moths that contained a similar number of species (small: average wingspan ≤ 2.75 cm; large: average wingspan > 2.75 cm).

D = − ∑1

2Pi


© 2006 The Authors 95Diversity and Distributions, 12, 91–99, Journal compilation © 2006 Blackwell Publishing Ltd

greater than or less than null expectation while another is not.

Our goal here was not to test for difference between two values

of α- or β-diversity that were both greater than (or less than)

expected by chance.

RESULTS

Of the 492 species for which we were able to obtain estimates of

body size, roughly equal numbers of species were considered

small and large, a pattern that did not vary much with season

(Table 2, Fig. 2). A far greater number of forest moth species

were considered dietary generalists compared to the number of

restricted feeders, however (Table 3). Body size did not appear to

differentially affect the spatial partitioning of α- and β-diversity

across sampling scales. In the early season, the partitions of

species richness for large and small-bodied species were virtually

identical (Fig. 3a). In contrast, using the Simpson index, small-

bodied species contributed 1.25–1.5 times as much observed

β-diversity among stands, sites and regions compared to larger-

bodied species (Table 2). Because the Simpson index is a measure

of dominance within communities, this result suggests that the

species-abundance distribution for smaller moths is less biased

by a few, highly abundant species. Indeed, the species-abundance

distribution of small moths more closely mirrors the log-series

distribution rather than the log-normal distribution characteristic

of larger-bodied species. In the late season, the observed partition

of species richness modestly reflected a non-significant bias

towards smaller species contributing a greater proportion of β-

diversity (Fig. 3b). There was no difference between partitions

for Simpson diversity (Table 2). For larger-bodied moths, α-

diversity was greater than that expected by chance, while observed

β-diversity was lower than expected among stands and greater

than expected among sites, and in the early season, between

ecoregions (P < 0.01 for each test of significance; Fig. 3a,b). For

smaller moths, β-diversity was significantly different from

random expectation (< than expected among stands, > than

expected among sites) only in the late season (P < 0.025, Fig. 3b).

Thus, the general trend to emerge from these diversity partitions

was that species richness of larger-bodied moths was more

similar than expected by chance among forest stands within a

single site, while sites and ecoregions contained significantly

greater levels of β-diversity than expected.

The effect of diet breadth on the contribution of species to either

α- or β-diversity was significantly different and less susceptible to

seasonal variation (Fig. 4a,b). In general, moth species that were

more restricted feeders had higher levels of β-diversity among

stand, sites and ecoregions but lower levels of species diversity

within stands (Table 3, Fig. 4). Furthermore, this trend was

observed for both species richness and Simpson diversity. In

contrast, generalists had higher levels of α-diversity, both in

Table 2 Additive partition of α- and β-diversity components across a hierarchically scaled inventory of forest Lepidoptera. Separate partitions were performed on small- and large-bodied moths in early and late sampling seasons

Season

Body

size

Diversity

component

Diversity measure

Species

richness

Simpson

diversity

Early Small Within stands (α1) 67.5 0.616

Among stands (β1) 55.3 0.035

Sites (β2) 47.2 0.015

Regions (β3) 38.0 0.036

Total 208.0 0.701

Large Within stands (α1) 54.1 0.826


Sites (β2) 38.9 0.003

Regions (β3) 32.4 0.038

Total 169.0 0.881

Late Small Within stands (α1) 52.0 0.939


Sites (β2) 49.2 0.009

Regions (β3) 44.6 0.006

Total 196.0 0.970

Large Within stands (α1) 51.9 0.948


Sites (β2) 39.0 0.007

Regions (β3) 35.4 0.011

Total 175.0 0.978

Table 3 Additive partition of α- and β-diversity components across a hierarchically scaled inventory of forest Lepidoptera. Separate partitions were performed on generalist and diet-restricted moths in early and late sampling seasons

Season

Diet

breadth

Diversity

component

Diversity measure

Species

richness

Simpson

diversity

Early Restricted Within stands (α1) 32.3 0.332


Sites (β2) 32.7 0.014

Regions (β3) 18.3 0.030

Total 114.0 0.395

Generalist Within stands (α1) 77.1 0.891


Sites (β2) 46.3 0.003

Regions (β3) 44.3 0.023

Total 228.0 0.928

Late Restricted Within stands (α1) 27.1 0.891


Sites (β2) 32.4 0.021

Regions (β3) 32.7 0.009

Total 125.0 0.950

Generalist Within stands (α1) 68.6 0.956


Sites (β2) 51.0 0.007

Regions (β3) 41.4 0.009

Total 222.0 0.981

K. S. Summerville et al.

96 © 2006 The AuthorsDiversity and Distributions, 12, 91–99, Journal compilation © 2006 Blackwell Publishing Ltd

terms of absolute species numbers (Table 3) and as a proportion

of the total species diversity (Fig. 4). Generalists also had higher

levels absolute measures of β-diversity at progressively large

spatial scales (Table 3). Simpson β-diversity for restricted feeders

at each spatial level was > 2 times that for generalists (Table 3),

suggesting that specialists were less variable species-abundance

distribution compared to generalists. Finally, observed α-diversity

components for the diet breadth comparisons were always

significantly greater than expected for dietary generalists and

significantly lower than expected by chance for dietary specialists

(Fig. 4a,b; P < 0.01). This result is consistent with expectation

when specialized species are aggregated in space, fewer species

will be present in a single forest stand, but community divergence

among stands is substantial (e.g. higher β-diversity). Generalists,

however, may appear more equitably distributed across spatial

scales, leading to higher α-diversity and lower β-diversity among

stands (Fig. 4a,b). Finally, β-diversity among sites and ecoregions

were greater than expected by chance for both specialists and

generalists, indicating that considerable species turnover occurs

regardless of diet breadth beyond the sampling scale of sites

(e.g. > 15 km) (Fig. 4a,b).

DISCUSSION

The principal goal of this study was to assess whether body size or

niche breadth influenced the partitioning of species diversity

across a hierarchy of spatial scales. We determined that differences

in body size did not greatly affect the partitioning of species

diversity, while differences in partitions based on diet breadth

were pronounced. This is consistent with the hypothesis that

host-plant distribution affects insect community composition at

local and regional scales (i.e. within and among forest stands),

and may exert a greater influence on species turnover and

changes in dominance within communities compared to dis-

persal limitation (at least as measured using body size). Similar

results are emerging from several recent studies of the effects of

habitat loss on insect communities (see Tscharntke et al., 2002;

Collinge et al., 2003; Summerville & Crist, 2004). Therefore,

occupancy of a given forest habitat may be due to the presence of

suitable vegetation and consequent aggregation of specialized

Figure 3 Percentage of total moth species richness and Simpson diversity explained by α and β components of regional diversity. Observed species diversity was partitioned for both large and small-bodied moth species within forest stands (α1), among forest stands (β1), among sites (β2) and between regions (β3). Separate diversity partitions were performed in (a) early and (b) late seasons. Asterisk indicates that the observed diversity at a level is significantly different than random expectation (P < 0.025).

Figure 4 Percentage of total moth species richness and Simpson diversity explained by α and β components of regional diversity. Observed species diversity was partitioned for moths considered to be either dietary generalists or restricted feeders within forest stands (α1), among forest stands (β1), among sites (β2) and between regions (β3). Separate diversity partitions were performed in (a) early and (b) late seasons. Asterisk indicates that the observed diversity at a level is significantly different than random expectation (P < 0.025).



herbivores rather than colonization dynamics per se (Brändle

et al., 2002).

Our test for body size, however, may have failed because, by

itself, body size is not a good predictor of geographical range in

forest moths (Blackburn & Gaston, 2003). Different life-history

traits may interact to determine species distribution (e.g. Gaston

et al., 2000; Ribeiro et al., 2003; but see Blackburn et al., 1997 for

a contrary case), so some very small species may actually have wide

distributions if they are highly vagile, have large fecundities, or

are general feeders (e.g. Chapman et al., 2002). Plutella xylostella

L. (Plutellidae), Argyroytaenia velutiana Clem. (Tortricidae) and

Dichomeris ligullela Hbn. (Gelechiidae) are small species that fit

such a description. Alternatively, large species may be patchily

distributed if necessary resources are themselves patchy, if adults

are relatively sedentary, or of the female is often flightless as in

some Lymantriidae and Geometridae (Nieminen, 1986; Doak,

2000). Nevertheless, studies at larger spatial scales than ours have

demonstrated positive correlations between body size, geographical

distribution, fecundity and dispersal ability in the Papilionoidea

(butterflies) (e.g. Gaston et al., 1998; Cowley et al., 2001). Thus,

chance effects and dispersal limitation may still have an important

role in determining lepidopteran community structure,

especially in highly disturbed or fragmented systems and at

larger spatial scales than considered in our study (Thomas &

Harrison, 1992; Doak, 2000; Thomas et al., 2000).

Flight ability and body size are also both known to be con-

strained by phylogeny in the Lepidoptera (Lindström et al., 1994),

and a posteriori test on our data revealed a significant effect of

family level phylogeny on body size ( PROC GLM, F = 24.44,

d.f. = 33, P < 0.01) despite some degree of overlap in range of sizes

represented by species in small-bodied families (e.g. Tortricidae,

Choristoneura spp.) and large-bodied families (e.g. Noctuidae,

Tarachidia spp.) (see also Koh et al., 2004). There was also broad

overlap in the range of body sizes for geometrid and noctuid

moths, but Geometridae are generally regarded as less vagile than

Noctuidae because of differences in body shape, wing morphology

and internal structure (Scoble, 1995). Because diversity parti-

tions for large- and small-bodied moth species were so similar,

however, we suggest that the lack of strict phylogenetic independ-

ence in our data does not detract from the central conclusion of

our analyses: diversity of large and small species appears to scale

the same way among forest stands, sites and ecoregions. Species

of Saturniidae, Sphingidae and Noctuidae (at the upper end of

our size range) were not necessarily more likely to have lower β-

diversity than species of Gelechiidae, Pyralidae or Tortricidae.

He and Legendre (2002) predicted and Veech et al. (2003)

demonstrated that α-diversity tends to be lower than expected by

chance and β-diversity tends to higher than expected by chance

when there is significant intraspecific aggregation among patches.

The significantly low levels of α-diversity within forest stands

for diet-restricted species are consistent with these patterns and

suggest that specialists tend to be aggregated in space. Further-

more, restricted feeders had higher levels of β-diversity among

stands than generalists, a pattern that would be expected if host

resources are patchy in distribution (Taylor, 1984; Ribeiro et al.,

2003; Summerville et al., 2003). Generalists, however, also had

significantly high levels of β-diversity at broad sampling scales.

We posit two mechanisms may contribute to high levels of β-

diversity in generalist moths. First, polyphagous moth species are

frequently characterized by substantial inter-annual variation in

population sizes compared to specialists (Rajmánek & Spitzer,

1982; Spitzer et al., 1984), so during poor years, populations may

not be readily detected in all sites. Second, Roy et al. (1998) noted

that the relationship between diet breadth and the geographical

distribution of British butterflies was complex and nonlinear.

Importantly, generalist species tended to have smaller ranges

than predicted by the presence of suitable host species, and this

would contribute to regional β-diversity if the same pattern

applied to forest moths.

As with body size, differences in diet breadth are widely known

to be constrained by phylogeny (e.g. Janz & Nylin, 1998), but our

data did not possess significant family level phylogenetic effects

on diet specialization ( PROC GLM, F = 1.43, d.f. = 28, P c.

0.10). This may not be a robust approach to resolving whether

phylogeny contributes to our diversity partitions because many

of the moth families listed in Table 1 contain both specialist and

generalist moths. Effects of phylogeny on diet breadth may be

most apparent at the level of tribe or genus. For example, within

the noctuid subfamily Catocalinae, species are reported to feed

on a wide variety of plants (Covell, 1983), but the genus Catocala

contains many species restricted to feeding on foliage from a

single family, or species, of host tree (Robinson et al., 2002). The

phylogeny of host plants within a site may also be important in

determining regional patterns of α- and β-diversity for Lepidop-

tera. Studies of broadscale geographical similarities among

insect communities emphasize the role of higher-level host plant

relatedness as a determinant of species turnover in phytophagous

insects (Ødegarrd et al., 2005). Thus, these studies posit an

alternative, evolutionary mechanism that is consistent with our

diversity partitions — communities of specialist species may

appear aggregated in space when phylogenetic similarity (and

thus plant chemistry and palatability — see Bernays & Chapman,

1994; Ødegarrd, 2005) among host plants is high locally and

divergent regionally.

In conclusion, in order to properly design a conservation plan

for this moth fauna, land managers will need to understand that

actions such as logging within a single stand will have a dis-

proportionately adverse affect on specialized species, which often

have even small home ranges than their host plants and can be

entirely lost from a region if highly aggregated at small spatial

scales (Quinn et al., 1997; Spitzer et al., 1997; Roy et al., 1998;

Summerville & Crist, 2002). Because β-diversity of specialists

among stands within a given forested site is relatively low but is

quite high at large sampling scales, preservation of single areas

such as watersheds in managed ecosystems will also fail to protect

more than a small subset of specialized species within an ecoregion

(see also Hamer et al., 1997). A preferred scheme for conservation

of forest Lepidoptera should therefore emphasize a stratified

approach to the design of preservation schemes. Such a design

might emphasize creating numerous forest preserves within

ecoregions even if each site is perceived to be relatively small (e.g.

c. 10 km2 in size).

K. S. Summerville et al.

98 © 2006 The AuthorsDiversity and Distributions, 12, 91–99, Journal compilation © 2006 Blackwell Publishing Ltd

ACKNOWLEDGEMENTS

The National Science Foundation (DEB-0235369), The Nature

Conservancy’s Ecosystem Research Program, the Ohio Board of

Regents Research Challenge Program, Sigma Xi and the Ohio

Biological Survey generously provided funding for this project.

C. Yeager, N. Anderson, C. Poling, B. Clarke, J. Kahn, J. Gering

and D. Golden helped with moth sampling and specimen

processing. E. Metzler, G. Balogh, L. Gibson, M. Nielsen and

C. Covell provided us with species identifications for taxa outside

our areas of expertise. D. Simberloff, L. Koh, C. Kremen, J. Shuey,

D. Courard-Hauri, M. Lewis and R. Steichen provided thoughtful

critiques of our work as this project developed. We are grateful to

the landowners and managers of the state parks, metroparks,

state forests and nature preserves used in this study, especially

P. Whan, C. Bedel, J. Watts, D. Karas and W. Oney. Without their

cooperation and assistance, this study would never have been

possible.

REFERENCES

Bernays, E.A. & Chapman, R.F. (1994) Host-plant selection by

phytophagous insects. Chapman and Hall, New York.

Blackburn, T.M. & Gaston, K.J. (1996) Spatial patterns in the

geographic range sizes of bird species in the New World.

Philosophical Transactions of the Royal Society of London. Series

B, Biological Sciences, 351, 897–912.

Blackburn, T.M. & Gaston, K.J. (2003) Macroecology: concepts

and consequences. Blackwell Science, Oxford.

Blackburn, T.M., Gaston, K.J. & Gregory, R.D. (1997) Abundance–

range size relationships in British birds: is unexplained variation

a product of life history? Ecography, 20, 466–474.

Bowman, J., Jaeger, J.A.G. & Fahrig, L. (2002) Dispersal distance

of mammals is proportional to home range size. Ecology, 83,

2044–2055.

Brändle, M., Öhlschläger, S. & Brandl, R. (2002) Range sizes

in butterflies: correlation across scales. Evolutionary Ecology

Research, 4, 993–1004.

Braun, E.L. (1961) Woody plants of Ohio. Haefner Press, New York.

Brown, J.H. (1995) Macroecology. University of Chicago Press,

Illinois.

Chapman, J.W., Reynolds, D.R., Smith, A.D., Riley, J.R.,

Pedgley, D.E. & Woiwod, I.P. (2002) High altitude migration of

the diamondback moth Plutella xylostella to the UK: a study

using radar, aerial netting, and ground trapping. Ecological

Entomology, 27, 641–650.

Collinge, S.K., Prudic, K.L. & Oliver, J.C. (2003) Effects of local

habitat characteristics and landscape context on grassland

butterfly diversity. Conservation Biology, 17, 178–187.

Covell, C.V.O. (1983) Field guide to eastern moths. Houghton-

Mifflin, New York.

Cowley, M.J.R., Thomas, C.D., Roy, D.B., Wilson, R.J.,

Leon-Cortes, J.L., Gutierrez, D., Bulman, C.R., Quinn, R.M.,

Moss, D. & Gaston, K.J. (2001) Density-distribution relation-

ships in British butterflies. I. The effect of mobility and spatial

scale. Journal of Animal Ecology, 70, 410–425.

Crist, T.O., Veech, J.A., Gering, J.C. & Summerville, K.S. (2003)

Partitioning species diversity across spatial scales: statistical

tests on components of insect diversity in deciduous forests.

American Naturalist, 162, 734–743.

Doak, P. (2000) Population consequences of restricted dispersal

for an insect herbivore in a subdivided habitat. Ecology, 81,

1828–1841.

Forbes, W.M.T. (1923) Lepidoptera of New York and neighboring

states. Part 1: Primitive forms. Cornell University Agricultural

Experiment Station, New York.

Gardezi, T. & da Silva, J. (1999) Diversity in relation to body size

in mammals: a comparative study. The American Naturalist,

153, 110–123.

Gaston, K.J. (1988) Patterns in local and regional dynamics of

moth populations. Oikos, 53, 49–57.

Gaston, K.J. & Reavey, D. (1989) Patterns in the life histories and

feeding strategies of British macrolepidoptera. Biological

Journal of the Linnean Society, 37, 367–382.

Gaston, K.J., Quinn, R.M., Blackburn, T.M. & Eversham, B.C.

(1998) Species-range size distributions in Britain. Ecography,

21, 361–370.

Gaston, K.J., Blackburn, T.M., Greenwood, J.J.D., Gregory, R.D.,

Quinn, R.M. & Lawton, J.H. (2000) Abundance-occupancy

relationships. Journal of Applied Ecology, 37, 39–60.

Gering, J.C., Crist, T.O. & Veech, J.A. (2002) Additive partitioning

of species diversity across multiple scales: implications for the

regional conservation of biodiversity. Conservation Biology, 17,

488–499.

Greller, A.M. (1988) Deciduous forest. North American terrestrial

vegetation (ed. by M.G. Barbour and W.D. Billings), pp. 287–

316. Cambridge University Press, London.

Hamer, K.C., Hill, J.K., Lace, L.A. & Langan, A.M. (1997) Ecological

and biogeographical effects of forest disturbance on tropical

butterflies on Sumba, Indonesia. Journal of Biogeography, 24,

67–75.

Hawkins, B.A. & Lawton, J.H. (1995) Latitudinal gradients in

butterfly body sizes: is there a general pattern? Oecologia, 102,

31–36.

He, F.L. & Legendre, P. (2002) Species diversity patterns derived

from species–area models. Ecology, 83, 1185–1198.

Hodges, R. (1999) The Gelechoidea. In: Handbook of zoology.

Volume IV Arthropoda: Insecta, Part 35: Lepidoptera, evolution,

systematics, and biogeography (ed. by Kristensen N.P.). W. de

Gruyter, Berlin.

Hodges, R.W., Dominick, T., Davis, D.R., Ferguson, D.C.,

Franclemont, J.G., Munroe, E.G. & Powell, J.A. (1983) Checklist

of Lepidoptera of America North of Mexico. Wedge Entomo-

logical Research Foundation, Washington, DC.

Hunter, A. (1991) Traits that distinguish outbreaking and non-

outbreaking Macrolepidoptera feeding on northern hardwood

trees. Oikos, 60, 275–282.

Janz, N. & Nylin, S. (1998) Butterflies and plants: a phylogenetic

study. Evolution, 52, 486–503.

Koh, L.P., Sodhi, N.S. & Brook, B.W. (2004) Ecological correlates

of extinction proneness in tropical butterflies. Conservation

Biology, 18, 1571–1578.



Krebs, C.J. (1999) Ecological methodology. Benjamin Cummings,

London.

Lande, R. (1996) Statistics and partitioning of species diversity

and similarity among multiple communities. Oikos, 76, 5–13.

LepÍ, J., Spitzer, K. & Jaros, J. (1998) Food plants, species

composition, and variability of the moth community in an

undisturbed forest. Oikos, 81, 538–548.

Lindström, J., Kaila, L. & Niemelä, P. (1994) Polyphagy and adult

body size in geometrid moths. Oecologia, 98, 130–132.

Longino, J.T., Coddington, J. & Colwell, R.K. (2002) The ant

fauna of a tropical rain forest: estimating species richness three

different ways. Ecology, 83, 689–703.

Magurran, A.E. (1988) Ecological diversity and its measurement.

Princeton University Press, Princeton.

Nieminen, M. (1986) Migration of a moth species in a network

of small islands. Oecologia, 108, 643–651.

Ødegarrd, F. (2005) Host specificity, alpha- and beta-diversity

of phytophagous beetles in two tropical forests of Panama.

Biodiversity and Conservation, in press.

Ødegarrd, F., Disrud, O.H. & Østbye, K. (2005) The importance

of plant relatedness for host utilization among phytophagous

insects. Ecology Letters, 8, 612–617.

Ouin, A. & Burel, F. (2002) Influence of herbaceous elements

on butterfly diversity in hedgerow agricultural landscapes.

Agriculture, Ecosystems, and Environment, 93, 45–53.

Quinn, R.M., Gaston, K.J. & Roy, D.B. (1997) Coincidence

between consumer and host occurrence: macrolepidoptera in

Britain. Ecological Entomology, 22, 197–208.

Rajmánek, M. & Spitzer, K. (1982) Bionomic strategies and long-

term fluctuations in abundance of Noctuidae (Lepidoptera).

Acta Entomolov Bohemoslov, 79, 81–96.

Redfearn, A. & Pimm, S.L. (1988) Population variability and

polyphagy in herbivorous insect communities. Ecological

Monographs, 58, 39–55.

Ribeiro, K.T., Codeco, C.T. & Fernandes, G.W. (2003) Local

and regional spatial distribution of an eruptive and a latent

herbivore insect species. Austral Ecology, 28, 99–108.

Robinson, G.S., Ackery, P.R., Kitching, I.J., Beccaloni, G.W. &

Hernández, L.M. (2002) Hostplants of the moth and butterfly

caterpillars of America North of Mexico. Memoirs of the

American Entomological Institute, 69, 1–824.

Roy, D.B., Quinn, R.M. & Gaston, K.J. (1998) Coincidence in the

distributions of butterflies and their foodplants. Ecography, 21,

279–289.

Scoble, M.J. (1995) The Lepidoptera: form, function, and diversity.

Oxford University Press, London.

Spitzer, K., JaroÍ, J., Havelka, J. & LepÍ, J. (1997) Effect of small-

scale disturbance on butterfly communities in an Indochinese

montane rainforest. Biological Conservation, 80, 9–15.

Spitzer, K., Rejmánek, M. & Soldán, T. (1984) The fecundity

and long-term variability in abundance of noctuid moths

(Lepidoptera, Noctuidae). Oecologia, 64, 91–93.

Summerville, K.S., Boulware, M.J., Veech, J.A. & Crist, T.O.

(2003) Spatial variation in species diversity and composition

of forest Lepidoptera in eastern deciduous forests of North

America. Conservation Biology, 17, 1045–1057.

Summerville, K.S. & Crist, T.O. (2002) Effects of timber harvest

on forest Lepidoptera: community, guild, and species responses.

Ecological Applications, 12, 820–835.

Summerville, K.S. & Crist, T.O. (2004) Contrasting effects of

habitat quantity and quality on moth communities in

fragmented landscapes. Ecography, 27, 3–12.

Summerville, K.S., Veech, J.A. & Crist, T.O. (2002) Does variation

in patch use among butterfly species contribute to nestedness

at fine spatial scales? Oikos, 97, 195–204.

Swihart, R.K., Gehring, T.M., Kolozsvary, M.B. & Nupp, T.E.

(2003) Responses of ‘resistant’ vertebrates to habitat loss and

fragmentation: the importance of niche breadth and range

boundaries. Diversity and Distributions, 9, 1–18.

Taylor, L.R. (1984) Assessing and interpreting the spatial distri-

bution on insect populations. Annual Review of Entomology,

29, 321–357.

The Nature Conservancy (1999) Conservation by design: a frame-

work for success. The Nature Conservancy, Virginia.

Thomas, C.D., Baguette, M. & Lewis, O.T. (2000) Butterfly

movement and conservation in patchy landscapes. Behavior

and conservation (ed. by L. Gosling and W. Sutherland).

Cambridge University Press, London.

Thomas, C.D. & Harrison, S. (1992) Spatial dynamics of a

patchily-distributed butterfly species. Journal of Animal

Ecology, 61, 437–446.

Thomas, A.W. & Thomas, G.M. (1994) Sampling strategies for

estimating moth species diversity using a light trap in a north-

eastern softwood forest. Journal of the Lepidopterists’ Society,

48, 85–105.

Tscharntke, T., Steffan-Dewenter, I., Kruess, A. & Thies, C.

(2002) Characteristics of insect populations on habitat

fragments: a mini-review. Ecological Research, 17, 229–

239.

Veech, J.A., Crist, T.O. & Summerville, K.S. (2003) Intraspecific

aggregation decreases local species diversity of arthropods.

Ecology, 84, 3376–3383.

Veech, J.A., Summerville, K.S., Crist, T.O. & Gering, J.C. (2002)

The additive partitioning of species diversity: recent revival of

an old idea. Oikos, 99, 3–9.

Wagner, H.H., Wildi, O. & Ewald, C.W. (2000) Additive parti-

tioning of plant species diversity in an agricultural mosaic

landscape. Landscape Ecology, 15, 219–227.

do body size and diet breadth affect partitioning of species diversity? a test with forest...

Documents